XEN cell lines are assumed to derive from the PrE lineage of blastocysts, which are cultured in the presence of FGF4 and heparin or ES medium with LIF (Kunath et al., 2005; Niakan et al., 2013). XEN cells share many
characteristics with PE, but few with VE. XEN cells contribute mostly to PE and rarely to VE after being injected into blastocysts (Kunath et al., 2005; Lin et al., 2016). XEN cells are a heterogeneous population and contain two distinct morphologies, a small round highly refractile morphology and a stellate epithelial-like morphology (Kunath et al., 2005). XEN cells can differentiate into cells with a VE identity in vitro either by BMP signaling, by Nodal and Cripto signaling (Kruithof-de Julio et al., 2011; Artus et al., 2012;
Paca et al., 2012), by laminin-mediated differentiation, or by high-density culture on gelatin-coated plates (Paca et al., 2012). When XEN cells
differentiate into cells with a VE identity, they will have increased expression of AFP (alpha-fetoprotein), E-cadherin, Ihh, and Ttr (Paca et al., 2012). XEN cells differentiated into cells with a VE identity contribute to VE and AVE after
being injected into blastocysts (Kruithof-de Julio et al., 2011). XEN cells express PrE specific genes, such as Gata6, Pdgfra, Sox17, Gata4, Sox7, Dab2, and Sparc (Kunath et al., 2005). How do such genes regulate XEN cells? The Erk pathway is critical for XEN cell derivation, and Grb2 is a critical adaptor in the Grb2-Erk pathway. Grb2 mutations block PrE cell lineage in blastocysts and block the expression of PrE-specific genes, such as Gata6, Pdgfra, Sox17, Gata4, and Sox7 (Chazaud et al., 2006). Fgfr2 is critical for the Grb2-Erk pathway, and inhibition of Fgfr2 by PD173074 or inhibition of MEK by PD0325901 has similar effects as Grb2 mutations (Nichols et al., 2009; Yamanaka et al., 2010). The active Erk pathway induces first Gata6 expression and then the other PrE specific genes. Gata6 mutations
completely block the subsequent expression of genes, such as Pdgfra, Sox17, Gata4, and Sox7 (Bessonnard et al., 2014; Schrode et al., 2014). Because Fgf4 can be replaced by FGF2, Fgf4 is not essential for the establishment of XEN cell lines; however, Fgfr2 could be essential for the establishment of XEN cell lines (Kang et al., 2013). Sall4 is essential for derivation of XEN cell lines. Sall4 seems to play a role as an activator of key lineage-defining genes in the ExEn (Lim et al. 2008). ExEn is hypomethylated when compared with embryonic tissue (Chapman et al., 1984; Monk et al., 1987; Gardner &
Davies,1992). XEN cells express low levels of H3K27me3 (Rugg-Gunn et al., 2010). Thus, hypomethylation of ExEn could enable the tissue to easily undergo differentiation or transdifferentiation.
PDGFRA and Sox17 are important to form the PrE cell lineage. Deletion of PDGFRA or Sox17 decreases the number of PrE cells (Artus et al., 2011, 2013). PDGFRA-mutant embryos can still develop to term but with severe defects, and they die soon after birth (Ogura et al., 1998). Sox17 mutant embryos cannot survive past the E8.0 stage (Artus et al., 2013). There are reports that PDGFRA and Sox17 are essential for deriving XEN cell lines (Artus et al., 2010; Niakan et al., 2010; Cho et al., 2012). We hypothesized that XEN cell lines can be derived from PDGFRA-mutant embryos and ES cells, because the remaining PrE still has the ability to support embryo
development. It is still not clear which genes regulate XEN cell derivation and maintenance.
Human XEN cell lines have not been derived from human blastocysts. This failure could be due to differences in the growth factors that support XEN cell progenitors between humans and mice (van Kuijk et al., 2012; Roode et al., 2012). Although overexpression of endodermal transcription factors results in expression of many endoderm markers, it is unclear what the molecular characteristics of XEN cells are (Séguin et al., 2008; Wamaitha et al., 2015).
A rat extra-embryonic endoderm precursor (XEN-P) cell line has been derived from rat blastocysts; the XEN-P cell line expresses Oct4 and SSEA1 at high levels, its growth is stimulated by LIF, and cells express Gata6 and Gata4.
XEN-P cells can contribute to ExEn after injection into rat blastocysts (Debeb et al., 2009). Zhong et al., 2018 reported the isolation of mouse primitive extraembryonic endoderm stem cell (pXEN) lines from mouse blastocysts, which express Oct4 and share characteristics with rat XEN-P cells. pXEN cells are highly similar to XEN cells by morphology, gene expression profile and lineage contribution. pXEN cells can convert into XEN-like cells, but not vice versa. pXEN cells are more representative than XEN cells of PrE of the blastocyst stage (Zhong et al., 2018). However, in this report, it is unclear if pXEN cells can contribute efficiently to VE-like PrE. In another recent report, mouse ES cells converted to XEN-like cells termed naïve extraembryonic endodermal (nEnd) cells, with characteristics close to blastocyst-stage ExEn precursors, by adding Activin A, LIF and Chir99021 in the culture medium (Anderson et al., 2017). Interestingly, rodent multipotent adult progenitor cell (MAPC) derived from bone marrow in rat multipotent adult progenitor cell (rMAPC) medium (Lo Nigro et al., 2012) express Oct4 and Rex1 but not Nanog and Sox2. However, MAPC express Gata4, Gata6, Sox7, Sox17, which are expressed in the PrE (Nichols et al., 2011) and in rat XEN-P cells (Debeb et al., 2009). Rat XEN-P cell lines derived from rat blastocysts in rMAPC medium, which resemble E3.5 nascent hypoblasts, were termed rat hypoblast stem cells (rHypoSC) (Lo Nigro et al., 2012). When green
fluorescent protein (GFP)-labeled rMAPC and rHyoSCs were aggregated with rat morulae, both types of cells contribute to ExEn (Lo Nigro et al., 2012).
Since bone marrow-derived MAPC have similar characteristics to rHypoSCs, some bone marrow cells may originate from PrE. It is unclear that XEN cells
can convert into bone marrow cells.
Mouse fibroblasts pass via a XEN-like state on their way to induced
pluripotent stem cell (iPSC) by chemical reprogramming (Zhao et al., 2015).
The same group reported that chemically induced pluripotent stem cells pass via a XEN-like stage to a 2C-like stage (early embryonic-like) to become iPS cells (Zhao et al., 2018). In the early or middle blastocyst stage, epiblast precursors can convert to PrE precursors spontaneously, and conversely, PrE precursors can convert to epiblast precursors as well (Grabarek et al., 2012).
However, in the E4.5 blastocyst, epiblast precursors show less plasticity than precursors of PrE, probably owing to differences in responsiveness to
extracellular signaling (Grabarek et al., 2012). Using single-cell resolution quantitative imaging, Xenopoulos et al., 2015 noted an irreversible
commitment to epiblast/PrE lineages in vivo and showed that rare cells from PrE can convert into epiblast, but not vice versa. In vitro, ES cells can convert into XEN cells spontaneously (Lin et al., 2017; Lo Nigro et al., 2017). However, there is no report that XEN cells can convert into ES cells spontaneously.
XEN cells could be converted into pluripotent stem cells by chemical induction (Zhao et al., 2015). It could be that XEN-like cells induced by chemicals are more similar to PrE-like cells and that the PrE-like cells convert to pluripotent stem cells.